A display device, using for example, liquid-crystal cells as display elements, includes a matrix of numerous pixels, and controls light intensity on a per pixel basis in response to image information to be displayed, thereby presenting a display on the pixels. An organic EL display employing organic EL elements is also driven in the same way.
However, the organic EL display, which is a self-emitting-type display using an emitting element as a display pixel, presents advantages of a high visibility of an image, compared with that provided by a liquid-crystal display, of requiring no backlight, and of a high response speed. The organic EL display is different from the liquid-crystal display in that the organic EL display is of a current control type while the liquid-crystal display is of a voltage control type. Specifically, luminance of the organic EL element is controlled by a current flowing therethrough.
A simple (passive) matrix method and an active-matrix method are available to drive the organic EL display in the same as a liquid-crystal display. Although being simple in structure, the former method cannot be used in a large-scale and high-definition display. For this reason, active-matrix displays are now actively being developed in which a current flowing through an emitting element in each pixel is controlled by an active element (a thin-film transistor (TFT)) arranged within a pixel.
FIG. 33 shows a pixel circuit (a circuit for a unit pixel) in a conventional active-matrix organic EL display (disclosed in U.S. Pat. No. 5,684,365 and Japanese Unexamined Patent Application Publication No. 8-234683).
Referring to FIG. 33, the conventional pixel circuit includes an organic EL element 101 with the anode thereof connected to a positive power source Vdd, a TFT 102 with the drain thereof connected to the cathode of the organic EL element 101 and the source thereof grounded, a capacitor 103 connected between the gate of the TFT 102 and ground, and a TFT 104 with the drain thereof connected to the gate of the TFT 102, with the source thereof connected to a data line 106, and with the gate thereof connected to a scanning line 105.
The organic EL element has a rectification feature, in many cases, so is sometimes referred to as an OLED (organic light emitting diode). Accordingly, the OLED is represented by a diode symbol in FIG. 33 and other figures. However, in the discussion that follows, rectification features are not a requirement.
The pixel circuit thus constructed operates as follows. Now, the scanning line 105 is in a selection state (at a high level, here) and the data line 106 is supplied with a writing potential Vw. The TFT 104 is turned on, charging or discharging the capacitor 103, and thereby the potential of the gate of the TFT 102 becomes the writing potential Vw. When the scanning line 105 is driven to a deselection potential (at a low level, here), the scanning line 105 is electrically disconnected from the TFT 102, but the gate voltage of the TFT 102 is reliably maintained by the capacitor 103.
A current flowing through the TFT 102 and the OLED 101 responds to a value of gate-source voltage Vgs of the TFT 102. The OLED 101 continuously emits light at a luminance level determined by the current value responsive to the gate-source voltage Vgs. In the following discussion, a “writing operation” refers to an operation to transfer luminance information, given to the data line 106, to within a pixel when the scanning line 105 is selected. As described above, in the pixel circuit shown in FIG. 33, once the writing operation is performed at the writing potential Vw, the OLED 101 continuously emits light at a constant luminance level.
Such pixel circuits (hereinafter also referred to as pixels) 111 are arranged in a matrix as shown in FIG. 34. A scanning line driving circuit 113 successively selects scanning lines 112-1 through 112-n while a data line driving circuit (a voltage driver) 114 of a voltage driving type writes data on data lines 115-1 through 115-m. The active-matrix display device (the organic EL display) is thus driven. The active-matrix display device here includes a matrix of n rows by m columns of pixels. In this case, the number of data lines is m, while the number of scanning lines is n.
In the passive-matrix display device, each emitting element emits light only at the moment it is selected. In the active-matrix display device, an emitting element continuously emits light even after the end of data writing. For this reason, the active-matrix display device outperforms the passive-matrix display device particularly in the field of large-scale and high-definition displays, because a low peak luminance and a low peak current of each light emitting element work in the active-matrix display device.
In the active-matrix organic EL display device, an insulated gate thin-film field-effect transistor (TFT) formed on a glass substrate is typically used as an active element. Since amorphous silicon or polysilicon used in the formation of the TFT generally suffers from poor crystallinity, and a poor controllability in the conductive mechanism thereof, a resulting TFT is subject to large variations in the characteristics thereof.
When the polysilicon TFT is formed on a relatively large-sized glass substrate, crystallization is usually performed using laser annealing subsequent to the formation of an amorphous silicon layer to control a thermal deformation of the glass substrate. However, it is difficult to uniformly irradiate a relatively large-sized glass substrate with laser energy, and the polysilicon suffers from localized variations in the crystallization state thereof. As a result, the threshold voltage Vth of the TFTs formed on the same substrate vary within a range of several hundreds of mV, in certain cases, 1V or more.
In this case, even if the same potential Vw is written on different pixels, the threshold value Vth of the TFT varies from pixel to pixel. The current Ids flowing through the OLED greatly varies from pixel to pixel, and the display device cannot be expected to present a high-quality image. Variations take place not only in the threshold value Vth but also in the mobility μ of the carrier.
The inventor of the present invention has proposed a current-programmed-type pixel circuit as shown in FIG. 35 to resolve the above problem (reference is made to International Publication No. WO01-06484).
A current-programmed-type pixel circuit includes an OLED 121 with the cathode thereof connected to a negative power source Vss, a TFT 122 with the drain thereof connected to the anode of the OLED 121, and with the source thereof connected to ground, which serves as a reference potential point, a capacitor 123 connected between the gate of the TFT 122 and ground, a TFT 124 with the gate thereof connected to the gate of the TFT 122 and with the source thereof grounded, a TFT 125 with the drain thereof connected to the drain of the TFT 124, with the source thereof connected to a data line 128, and with the gate thereof connected to a scanning line 127, and a TFT 126 with the drain thereof connected to each of the gates of the TFT 122 and the TFT 124, with the source thereof connected to each of the drains of the TFT 124 and the TFT 125, and with the gate thereof connected to the scanning line 127.
In this circuit, the TFT 122 and the TFT 124 are PMOS field-effect transistors, and the TFT 125 and the TFT 126 are NMOS type. FIGS. 36A to 36C are timing diagrams of the pixel circuit in the driving operation thereof.
The pixel circuit shown in FIG. 35 is different from that shown in FIG. 33. Luminance data is given in the form of voltage in the pixel circuit shown in FIG. 33, while the same data is given in the form of current in the pixel circuit shown in FIG. 35. The operation of the circuit shown in FIG. 35 will now be discussed.
To write the luminance information, the scanning line 127 is set to a selection state and a current Iw corresponding to the luminance information flows through the data line 128. The current Iw flows through the TFT 124 via the TFT 125. The gate-source voltage generated between the gate and the source of the TFT 124 is referred to as Vgs. During the writing operation, the TFT 124 operates in the saturation region thereof because the TFT 126 shorts the gate and the drain of the TFT 124.
The following well-known equation of the MOS transistor holds.Iw=μ1Cox1W1/L1/2(Vgs−Vth1)2  (1)
In equation (1), Vth1 is a threshold value of the TFT 124, μ1 is the mobility of the carrier, Cox1 is the gate capacitance per unit area, W1 is the channel width, and L1 is the channel length.
A current flowing through the OLED 121 is referred to as Idrv, the current Idrv is controlled the value by the TFT 122 connected in series with the OLED 121. In the pixel circuit shown in FIG. 35, the gate-source voltage of the TFT 122 agrees with Vgs in the equation (1). On the assumption that the TFT 122 operates in the saturation region thereof, the following equation (2) holds.Idrv=μ2Cox2W2/L2/2(Vgs−Vth2)2  (2)
The condition under which the MOS transistor operates in the saturation region thereof is expressed by the following equation (3).|Vds|>|Vgs−Vth|  (3)
The symbols in the equations (2) and (3) are identical to those used in the equation (1). Since the TFT 124 and the TFT 122 are formed closely in a small area within the pixel, in practice, μ1=μ2, Cox1=Cox2, and Vth1=Vth2. From the equations (1) and (2),Idrv/Iw=(W2/W1)/(L2/L1)  (4)
Even if the mobility μ of the carrier, the gate capacitance Cox per unit area, and the threshold value Vth are varied within a panel, or from panel to panel, the luminance of the OLED 121 is precisely controlled because the current Idrv flowing through the OLED 121 is accurately proportional to the writing current Iw. For example, if the transistors are designed with the conditions of W2=W1 and L2=L1 satisfied, Idrv/Iw=1. Specifically, the writing current Iw equals the current Idrv flowing through the OLED 121 regardless of variations in the TFT characteristics.
In the active-matrix display device, the writing of the luminance data to each pixel is basically performed on a scanning line by scanning line basis. For example, in a liquid-crystal display using amorphous silicon TFTs, the writing of the luminance data is performed on the pixels arranged on a selected scanning line at a time basis. The writing on a per scanning line basis is now referred to a line-by-line writing operation.
In the display device working on a line at a time writing operation, the data line driver is manufactured using a typical monolithic semiconductor technology in a manufacturing process different from the manufacturing process of the pixel circuit (TFT) in the display panel. A data line driving circuit having reliable characteristics is thus easily manufactured. On the other hand, since it is necessary to have a plurality of data line drivers, the number of which is equal to the number of data lines in the display device, the entire system becomes bulky in size and costly. To manufacture a display device having a large number of pixels or pixels arranged in a narrow pitch, the number of lines and connections of a display panel with the drivers external to the panel become large. The effort to develop a large-scale and high-definition display device is subject to a limitation in terms of the reliability of the connections and the wiring pitch.
The “drivers external to the panel” are literally arranged outside the display panel (the glass substrate), and are occasionally connected to the panel using a flexible cable. The drivers external to the panel are sometimes mounted on the panel (the glass substrate) using the TAB (Tape Automated Bonding) technology. The phrase “drivers external to the panel” is and will be used in the context of the above two arrangements.
With its high transistor driving performance, the liquid-crystal display using the polysilicon TFT writes data on a single pixel for a short period of time, and a dot-by-dot writing operation is typically adopted. FIG. 37 shows the construction of a display device working on a dot-by-dot writing operation and FIGS. 38A to 38F are timing diagrams of the display device. Note that in FIG. 37, the same parts as those of FIG. 34 are indicated by the same symbols as those of FIG. 34.
Referring to FIG. 37, horizontal switches HSW1-SHWm are respectively connected between the ends of data lines 115-1 through 115-m and a signal input line 116. The horizontal switches HSW1-HSWm are turned on and off by selection pulses we1-wem that are successively output from a horizontal scanner (HSCAN) 117. The horizontal switches HSW1-HSWm and the horizontal scanner 117 are formed of TFTs, and are manufactured in the same manufacturing process as that of a pixel circuit 111.
The horizontal scanner 117 receives a horizontal start pulse hsp and a horizontal clock hck. Referring to FIGS. 38A to 38F, subsequent to the input of the horizontal start pulse hsp, the horizontal scanner 117 successively generates the selection pulses we1-wem to select the horizontal switches HSW1-HSWm, in response to the transition of the horizontal clock hck (the rising edge or the falling edge of the horizontal clock hck).
Each of the horizontal switches HSW1-HSWm becomes conductive when the corresponding one of the selection pulses we1-wem is fed, thereby transferring image data (a voltage value) sin to each of the data lines 115-1 through 115-m through the signal input line 116. In this way, the writing of the data on the pixels of the scanning line selected by the scanning line driving circuit 113 is performed on a dot-by-dot basis. The voltage given to the data lines 115-1 through 115-m is held by a capacitive component such as a stray capacity of each of the data lines 115-1 through 115-m even after the horizontal switches HSW1-HSWm becomes non-conductive.
When m clocks of the horizontal clock hck are fed, the data is written on all pixels on the selected scanning line. Since the display device working on a dot-by-dot basis uses the single signal input line 116 on a time sharing manner, the number of connection points between the display panel and the data line drivers (a circuit for feeding the image data sin) external to the display panel is small in number, and the number of the external drivers is accordingly small.
When the current-programmed-type pixel circuit shown in FIG. 35 is adopted as the pixel circuit, however, it is impossible to normally write the data on the pixels 111 in the display device shown in FIG. 37. The reason for this will be discussed.
When the signal input line 116 is driven by a current source with a particular horizontal switch HSW being selected and conductive in FIG. 37, a normal current writing is performed on a pixel on a data line of the selected horizontal switch HSW. When the current writing starts on another data line with the horizontal clock hck input to the horizontal scanner 117 thereafter, the horizontal switch HSW, which was selected until then, becomes conductive at the moment of writing. The current flowing into the corresponding data line becomes zero.
To perform the normal writing, a predetermined writing current needs to be fed to all pixels on the scanning line when the scanning lines are switched from the selection state to the deselection state thereof. In other words, when the current-programmed-type pixel circuit is adopted, the data writing on the pixels needs to be performed on a line-by-line basis. Referring to FIG. 39, a data line driver 118 arranged external to the display panel needs to be used to concurrently write the data onto the pixels on the selected scanning line.
The circuit shown in FIG. 39 is essentially identical in construction to the circuit of a line-by-line driving method shown in FIG. 34. As a result, the circuit shown in FIG. 39 has the problem that the number of current drivers CD1-CDm forming the data line driving circuit 118 and the number of connection points between the current drivers and the display panel increase.